A Surveillance Epidemiology and End Results (SEER) database review of epidemiology and outcome of ampullary cancer compared to cholangiocarcinoma.

e15147 Background: Ampullary adenocarcinoma (AC) is a rare form of cancer with sparse epidemiological and survival data in literature. We aimed to study the epidemiology and survival outcome of AC compared to Cholangiocarcinoma (CC) by reviewing the SEER database. Methods: The SEER database(version 8.1.2) was reviewed for patients (pts) with histologically confirmed AC and CC between 1973-2007. Variables included were: age, race, gender, surgical resection of primary tumor, SEER adjusted stage and overall survival (OS). Primary outcome was 3 year OS. Data was analyzed using Chi square and Kaplan-Meier method using z-test. Results: A total of 3,179 pts with AC and 5273 with CC were included in SEER database between 1973 and 2007. The mean age in AC was higher 67.5 compared to 64.7 years in CC(P value=<0.0001). AC had higher white to black ratio (13:1) compared to CC (10:1) (P value=0.0016). Male to female ratio was similar in both groups. Stage frequency at presentation in AC were localized (23%), regional (65%) and distant (12%) where in CC they were 22%, 36% and 42%, respectively (P value=<0.0001). Surgery rate was higher in AC compared to CC in localized (64 versus 36%, P value=<0.0001), regional (82 versus 35%, P value=<0.0001) and distant stages (22% versus 13%, P value=<0.0001). When compared stage to stage for all pts, AC had better 3 year OS in localized (43% versus 21%), regional (43% versus 12%) and distant disease (3.9% versus 2.6%) as compared to CC. In pts who had surgical treatment, AC still had better 3 year OS in localized (68 versus 59%, P-value=<0.0001), regional (44 versus 28%, P-value=<0.0001) and distant disease (12 versus 9%, P-value=0.45). Conclusions: AC as compared with CC, has statistical significant difference in age and race distribution. When compared stage to stage, AC has a better 3 year OS even after adjusting for potential confounders including the higher surgical resection rate. This difference appears to be statistically insignificant in distant stages.